FR2796507A1 - METHOD FOR PROVIDING A PULSE WIDTH MODULATED SIGNAL - Google Patents

Abstract

A method for providing a pulse width modulated signal having a high level signal of a first width and a low level signal of a second width. In a first operating state, the width of the high level signal is transmitted and / or continues to be processed and in a second operating state, the processing of the width of the low level signal is transmitted and / or continued according to a preset pulse rate.

Description

The present invention relates to a method and a

  device for generating a signal modulated in width of

  pulse having a high level signal from a first lar-

  geur and a low level signal of a second width.

  In many applications, for predefi-

  set a setpoint, for example a setpoint

  sign of current in a power stage, using a microcontroller, an analog value must be supplied. A

  such analog value is subject to specific requirements

  nant the quality of smoothing or filtering which must be as good as possible and simultaneously present a

  very great dynamics. As another example, there are the re-

  digital regulators requiring digital converters

  ques / analog to give a calculated quantity of

  adjustment or actuation to an analog type regulation

  than. For this, pulse width modulators are frequently used providing a pulse width proportional to a predetermined numerical value; this width then corresponds to an analog voltage after a low-pass filtering. The advantage of such pulse width modulators is their very low cost. It is for example possible to use a microcomputer with

  a pulse width output or to integrate the electronic-

  only digital, necessary in an ASIC card. The additional costs for a digital to analog converter often represent only the cost of an RC element which

  can be used as a simple low pass filter. Such conver-

  digital-analog weavers also have a li-

  remarkable age and monotony.

  In pulse width modulators ac-

  used as digital / analog converters

  However, it was found that their dynamics were re-

  latently slow and that they presented a spectrum of

  relatively low emission frequency.

  To remedy these drawbacks, the publica-

  tion Automatisierungstechnik 41 (1993) 11, pages 428-432, proposed to use pulse number modulators

  instead of pulse width modulators. The avan-

  the use of pulse number modulators

  sions indicated in this document is that compared to a pulse width modulation, for the same resolution over time, we have much higher emission frequency spectra and we can vary the set value more quickly. In addition, compared to pulse width modulation, smaller and thus less expensive filter stages can be used. The downside of the

  however, modulation of the number of pulses is necessary.

  ter a particular circuit in particular with calculation algorithms. This increases the cost of investment and

  operation of a pulse number modulator.

  There are certainly in micro-computers

  currently available, intelligent peripheral units

  gentes or programmable that can generate pulse width modulated (PWM) signals at very high resolution (e.g. 20 MHz clock, 50 ns resolution for separate events), but frequently these units are limited in that the distance between two successive signal flanks is limited downwards, for example to 10 ps at

  because of recharge or calculation times.

  The present invention aims to develop a pulse width modulated signal having a better resolution compared to that of known signals and a larger value range, the frequency of which is as high as possible to allow a filter stage in

  downstream with a limit frequency as high as possible.

  In particular, the transmission frequency must be able to

itself be changed.

  To this end, the invention relates to a method of the type defined above, characterized in that the width of the high level signal in a first state is provided and continued to be processed as a function of a predetermined pulse rate. and the signal width of

  low level in a second operating state.

  The invention also relates to a device of the

  type defined above, characterized by means for com-

  mutate between a first operating state in which the signal width is generated and / or continues to be processed

  high level based on a predetermined pulse rate-

  mined, and a second operating state in which the signal width is transmitted and / or continues to be processed

  low level according to a predefined pulse rate.

  According to the invention, a pulse width modulated signal is generated having a better resolution than

  that of known signals within the range of

  their desired and whose frequency is higher compared to that of known solutions, which allows for example a downstream filter stage having a

higher limit frequency.

  According to other advantageous features of the invention:

  * for a pulse rate of less than 50% between the

  high level signal and low level signal,

  nit a minimum signal level width allowed

  with respect to the high level signal and we send the if-

  low level depending on the preset pulse rate

  neither, and for a pulse rate of more than 50%, we

  sets a signal level width, minimum, auto-

  compared to the low level signal and the width of the high level signal is emitted as a predetermined pulse rate, * a minimum pulse rate for example of 10% and / or a maximum pulse rate for example is predefined

90%.

  For a pulse rate of less than 50% in

  be the high level signal and the low level signal of the pulse width modulated signal, it is advantageous to define a minimum signal level width allowed for the high level signal, and the low level signal is given as signal modulated in pulse width or is processed; for a pulse rate greater than 50%; an allowed nominal signal level range is defined for the low level signal, and the high level signal is given or processed as a pulse width modulated signal. The pulse rate is the ratio of the durations of the high level signal and the low level signal; a pulse rate of less than 50% means that the low level signal lasts more than half of the total period

of the signal.

  Advantageously, a rate is predefined

  minimum pulse, for example 10% and an

  maximum impulses, for example 90%. These means avoid emission frequencies going towards infinity, which one would have for pulse rates close to zero or%. It is also advantageous for the switching between the different operating states to be carried out with

hysteresis.

  The present invention will be described below in more detail with the aid of the appended drawings in which: * FIG. 1 is a schematic view of different states of a signal modulated in pulse width, * FIG. 2 shows schematically a typical characteristic of a pulse width modulator according to the invention, the output frequency of which is represented as a function of the pulse rate, * FIG. 3 shows a flowchart for the presentation

of the process of the invention.

  In pulse modulators, due to recharging or calculation times, we are required to respect a

  minimum interval between two successive sides of signals.

  A corresponding duration T_ min is given in Figure la

  for a characteristic signal, modulated in pulse width

  if we. The rate of TV pulses between the duration TH of the high level signal and the duration TL of the low level signal is in this case greater than 50%. The sum of the two durations or widths TH, TL results from the period of the output signal T-P of the signal modulated in pulse width. The duration of the low level signal TL corresponds to the minimum duration to respect Tmin between two successive signal flanks. By setting the duration of the low level signal TL to the value Tmin, the frequency fp of the signal modulated in pulse width can be maximized (the inverse of the period TP)

  without influencing the possibility of processing the mo-

  pulse width. Under these conditions, the duration TH corresponding to a predetermined pulse rate is transmitted as a signal. This duration is greater than

  Tmin (TV> 50%) and is not critical.

  By decreasing the duration TH of the high level signal, one can continue to decrease the period TP of the pulse width modulated signal until the duration of the level signal TH corresponds to the minimum distance T_ min between two neighboring flanks of signals. This siguation

  is shown in Figure lb. We see that under these conditions

  the period TP of the pulse width modulated signal is minimum; this means that its frequency f is at fmax. The TV pulse rate is in this case equal

at 50%.

  To further decrease the pulse rate,

  the duration of the level signal TL is increased; and we challenge

  nit Tmin as high, minimum level duration (figure lc).

  the duration TL is supplied as a signal, for example in function

  tone of the pulse rate to be predetermined. The duration TL is greater than Tmin (TV <50%) and therefore it is not critical. It is essential that when a pulse rate of 50% is reached, it is possible to switch between the two states "emission of the high level signal corresponding to a predetermined pulse rate" and "emission of the signal

  low level corresponding to a predefined pulse rate

  The switching between the two states can be carried out using suitable control means, for example a microcontroller associated with a pulse width modulator.

  weak heterestesis. Figure 2 shows the character curve

  corresponding statistics that the output frequency f of the pulse width modulated signal is plotted as a function of the TV pulse rate. We see that for an impulse rate of 50%, there is a continuous passage of the

  characteristic curve, but there is no jump. The brother

  quence f, emitted, represented in FIG. 2 is maximum for a TV pulse rate equal to 50%. Across the beach

  interesting pulse rates to be predetermined (in line

  suppressing for example the pulse rate from 10% to 90%), there is always an output signal with an output frequency a optimally adapted, i.e. as high

  as possible. Using such an output frequency signal

  tie, optimized, we can realize the filter stage downstream

  with a correspondingly raised limit frequency

dante.

  The method described makes it possible, using

  conventional devices, to achieve the highest possible frequency (and thus use the fewest filters possible

  as dynamic as possible). Continue

  vre optimization is only possible with a sup-

  additional (for example with modulation of the number of pulses

  sions, or complex filters, or higher frequencies). The method according to the invention will be described a

  again using the flowchart in Figure 3.

  In a first step 101, with control means

  appropriate not shown it is determined whether the pulse rate

  sions is greater than or equal to 50%. If so, we continue with a derivation between a step 102 in which we define Tmin as the minimum authorized width

  or duration of the low level signal belonging to the low signal

  pulse width. In step 103, we provide

  as signal the duration of the corresponding high level signal

  at the rate of pulses to be predefined.

  If on the contrary in step 101, it is noted that the pulse rate is less than 50%, there will be a bypass until step 104 in which T min is emitted

  as the minimum permitted width or duration of the ni-

  high calf. Then in step 105, depending on the pulse rate

  sions to be predetermined, we have the emission of the duration of the

low level signal.

Claims (5)

R E V E N D I C A T 1 0 N S   1) Method for generating a signal modulated in width of im   pulse having a high level signal from a first lar-   geur and a low level signal of a second width, characterized in that the width of the high level signal in a first operating state and the width   of the low level signal in a second operating state is lying.   2) Method according to claim 1, characterized in that for a pulse rate of less than 50% between the high level signal and the low level signal, a   minimum signal level width allowed by comparison   port to the high level signal and we send the signal of ni-   low calf according to the predefined pulse rate, and for a pulse rate of more than 50%, a width is defined   signal level, minimum, allowed with respect to   low level and the width of the ni-   high calf as a predetermined pulse rate.   3) Method according to claim 2, characterized in that a minimum pulse rate is predefined for example of 10%   and / or a maximum pulse rate for example of 90%.   4) Device for transmitting a pulse width modulated signal having a high level signal of a first width and a low level signal of a second width, characterized by   means for switching between a first operating state   in which one generates and / or one continues to process the width of the signal of the high level according to a rate   of predetermined pulses, and a second operating state   ment in which it transmits and / or continues to process the width of the low level signal according to a predefined pulse rate. 5) Device according to claim 4, characterized in that the switching between the respective operating states   is done by applying hysteresis.